Performance of Thin-Film Lithium Energy Cells under Uniaxial
Pressure**
By Tony Pereira,* Roberto Scaffaro, Zhanhu Guo, Simon Nieh, Jeff Arias and H. Thomas Hahn
Lithium batteries have attracted much interest in recent
years due to their superior performance. They consist of a
lithium-ion intercalation cathode and a lithium-metal anode
separated by a lithium-ion conducting electrolyte. Over the
past decades, much effort has been spent to improve the electrodes and electrolytes for better performance in terms of
energy storage, robustness and durability.[1,2] One of the most
important goals in this direction has been the design of thinfilm lithium batteries since the early 1990’s.[3–6] However, real
interest arose only with the development of the lithium phosphorus oxynitride (LiPON) solid electrolyte.[7–11] These batteries with thicknesses as thin as 10 lm offer an excellent combination of power and energy density. Such unique physical
properties allow their usage in multifunctional energy-storage structures which are not only load bearing but also energy storing. Typical examples of multifunctional applications
are electrically propelled unmanned aerial vehicles (UAVs)
and micro air vehicles (MAVs), in which the integration of
the batteries in the structure is crucial for their full realization.[12,13] The advent of thin-film batteries opens the possibility to use local autonomous power in new applications such
as micro-electro-mechanical systems (MEMS), tire pressure
sensors, and broadband antenna design.[22]
–
[*] Dr. T. Pereira, Dr. Z. Guo
Mechanical and Aerospace Engineering Department
University of California, Los Angeles, CA 90095
E-mail: [email protected]
Dr. R. Scaffaro
Dipartimento di Ingegneria Chimica dei Processi e dei
Materiali
University of Palermo, Italy
Dr. S. Nieh, Dr. J. Arias
Front Edge Technologies Inc.
Baldwin Park, CA
Prof. H. T. Hahn
Materials Science and Engineering Department
University of California, Los Angeles, CA 90095
E-mail: [email protected]
[**] This work was carried out at the Multifunctional Composites
Laboratory (MCL), University of California Los Angeles under
the Air Force Office of Scientific Research MURI Grant
FA9550-05-1-0138 to University of Washington, managed by
Dr. B. Les Lee.
ADVANCED ENGINEERING MATERIALS 2008, 10, No. 4
In view of such applications, and to increase the mechanical reliability of these systems, it is worth knowing the relations between the electrical performance, in particular charge
and discharge cycles, and other physical parameters such as
temperature, flexural deformation and uniformly distributed
pressure over their entire surface area.
The effect of temperature has been widely studied, together with some solutions proposed to increase the thermal stability of the batteries.[14–17] Above a certain temperature, reactions occurring in the electrodes rapidly lead to the failure of
the energy cells. A number of studies have been carried out
to find out the relations between the electrical behaviour and
the temperature[14–16] including the effect of thermal stabilizers[17] and the modelling of the phenomena.[18] The all-solidstate thin-film lithium energy cells used in this study have an
upper operating temperature limit of 150 °C claimed by the
manufacturer, albeit operation at higher temperatures will
result in shorter lifetimes. While the study of the effect of temperature exists, no literature is available to correlate the mechanical loading of uniformly distributed uniaxial pressure
with the electrical charge/discharge performance of these
energy cells.
The aim of this work series is to find the relations between
the electrical performance of all-solid-state thin-film lithium
energy cells when subjected to mechanical stresses. In the
first part of this work,[19] we analyzed the effect of flexural
deformations on the charge/discharge characteristics of these
thin-film energy cells. In this present work, the effect of uniformly distributed uniaxial surface pressure on the electrical
performance of the energy cells was investigated. The knowledge of the relations between pressure and electrical properties of the energy cell is necessary to fabricate the laminated
composites with these devices embedded. The fabrication of
laminated composites requires a constant pressure of about
550 kPa in an autoclave with curing times of one to several
hours. Although flexural bending tests and pressure tests are
similar in terms of their mechanical nature, they do not give
the same response in the energy cells because of the different
solicitations on the several components of the energy cell.
Hence, it is required to study these solicitations individually
and separately.
Experimental Procedure
All energy cells in this study were provided by Front Edge
Technologies (Baldwin Park, CA), with square dimensions of
about 25 mm on the side and average mass of (0.175 ± 0.025)
© 2008 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
393
COMMUNICATIONS
DOI: 10.1002/adem.200700214
COMMUNICATIONS
Pereira et al./Performance of Thin-Film Lithium Energy Cells under Uniaxial Pressure
g from a sample pool of 10 specimens. These energy cells are
currently in commercial production, and have a nominal voltage of 3.6 VDC (20). The all solid state thin-film lithium energy cells have a lifetime of 1,000 charge/discharge cycles of up
to 100 % depth-of-discharge, and a specific energy of about
200 Wh/kg without considering the packaging mass. In its
present configuration, accounting for the packaging mass,
yields a capacity of about 2 mAh/g, corresponding to an
energy density of about 7.2 Wh/kg. The packaging for the
configuration used in this study consists of two equal
top and bottom mica substrates of about 50 lm thickness
each. The mica used is muscovite, a phyllosilicate mineral
of aluminium and potassium with chemical formula KAl2(AlSi3O10)(F,OH)2 having a highly perfect basal cleavage that
yields remarkably thin laminae, often highly elastic. The mica
substrates are held together all around by a 40 lm polymer
layer of Surlyn (Dupont) sealant (see Fig. 5). While Surlyn is
permeable to ambient oxygen and water vapor, its permeability is very low. The Surlyn low permeability and the energy
cell high aspect ratio (cross section thickness vs. length of the
sealant) at the edges justifiy its use as a sealant in thin-film Li
batteries. This sealing strategy gives a maximum 2-year battery lifetime. The geometric profile of specimen 0606015 is
shown in Figure 1. Circles refer to the thickness in the area of
active components of the energy cell, i.e., anode, cathode,
electrolyte and the mica substrates. Boxes refer to the thickness at the edge all around the energy cell where there are no
active components, i.e., only the mica substrate and the sealant exist.
Two energy cells were initially subjected to three initial
consecutive charge/discharge cycles under no uniaxial pres-
Anode
Cathode
0.040
0.132
0.133
0.139
0.140
0.040
0.132
3
0.134
0.142
0.139
0.138
0.131
0.131
0.138
0.132
0.135
0.140
0.141
0.140
0.133
3
0.137
0.132
0.132
0.134
0.134
3
25
3
Fig. 1. Geometry and thickness profile (dimensions in mm). Circles refer to the thickness in the area of active components of the energy cell. Boxes refer to the thickness at
the edge all around the energy cell.
394
Micrometer
Deflection
http://www.aem-journal.com
Keithley
SourceMeter
Plotting &
Graphing
PC/MDS C
Control
GPIB
Data
Matlab - Discrete Data
Integration & Analysis
Program Control
Data
Logging
Fig. 2. Data acquisition and control flow diagram.
sure to establish a performance baseline for reference and
comparison. The charging was done at a constant voltage of
4.2 VDC and a limiting current draw of 1 mA using a Keithley SourceMeter 2400 under GPIB control of a MDS C program written specifically for this purpose. The data acquisition and analysis flowchart is shown in Figure 2. The real
time discrete current and voltage were measured at one-second intervals and logged to a text data file with cycle start/
end times and time sequence for each data set. Each specimen
was considered to be fully charged when the charge current
fell below 50 lA. Each charge cycle was immediately followed by a discharge at 1 mA constant sink current. The
device was considered to be fully discharged when the voltage dropped to a value of 3.0 VDC. The energy cell was held
flat, that is, without flexural deformations, for the entire set of
tests done in this study. The charge/discharge current of
1 mA used in this work corresponds to an average 2.5 C discharge ratio (1 mA/0.4 mAh), which is more severe than the
commonly used nominal discharge ratio of 1 C found in the
literature.[23] However, the former value is within the normal
operating parameters for this type of batteries, which can accept charge/discharge ratios higher than what are normally
expected without suffering damage.
The discharge is used to define the energy cell capacity C
in units of mAh. The discrete data acquired was taken to
Matlab and integrated using the trapezoidal method to calculate the energy cell capacity, with the total charge/discharge
adjusted for time accuracy. The energy cell capacity is
defined as follows:
Z
0.133
25
Sample in
Fixture
Cˆ
s
idt
…1†
0
where s is the upper limit of integration corresponding to the
time during charge at which the current i(t) = 50 lA is
reached, or the time at which during discharge the voltage
v(t) = 3.0 VDC is reached. The first three consecutive charge/
discharge cycles under no physical solicitations were used to
establish the characteristic baseline behaviour of each energy
cell.
A fixture was designed and fabricated by the author using
an insulating machinable wax of appropriate strength on a
micro-machining centre. Machining tolerances for the fixture
were kept at very low levels, to within less than about 50 microns, measured with Mitutoyo digital calipers. The schematic diagram of the fixture is shown in Figure 3, with the
actual pictures in Figures 4(a), 4(b) and 4(c). A cavity was
© 2008 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
ADVANCED ENGINEERING MATERIALS 2008, 10, No. 4
Pereira et al./Performance of Thin-Film Lithium Energy Cells under Uniaxial Pressure
3 mm
Rubber
Thin-Film
Energy
N.A.
Anode
Cathode
Fig. 3. Schematic of pressure test fixture.
1
2
3
Fig. 5. Cross-sectional schematic of energy cell. Scale is approximate. 1. Sealant 40 lm
(mica substrate bends slightly at the edge); 2. Active Component Layer: anode (Li-metal 2 lm thick, LiPON electrolyte 2 lm thick, lithium cobalt oxide cathode 6 lm thick,
respectively); 3. Mica substrate 50 lm thick.
ter. The conductive rubber pad between the energy cell electrode and the alligator clips may contribute to some energy
dissipation. However, due to the very low thin-film thickness
of the energy cell electrodes (100 nm) and the dynamic character of the experiment, the rubber foam contact was found to
be both more durable and reliable.
A tensile test was performed on another separate energy
cell to determine its mechanical properties. The Young’s
Modulus was found to be 107 GPa, and the failure strength
was found to be 80 MPa. The measured tensile properties of
the energy cell are attributed mostly to the mica (muscovite)
substrate.[19] The cross-sectional profile reveals that the active
components (anode, electrolyte and cathode) are very thin,
∼ 10 lm, relative to the total thickness of the two mica substrates, ∼ 100 lm, Figure 5.
Results and Discussion
Fig. 4. Rapid Prototype pressure test fixture; a) Open, showing electrical contacts and
rubber cushioning; b) Closed, assembled; c) Installed on a mechanical testing machine.
formed to accommodate the energy cell specimen, with channel conduits for the electrode terminal connections, and sufficient height to cushion both the top and bottom sides of the
energy cell with about 3 mm of a soft rubber material. Electrical contacts to the energy cell were established by using a
conductive rubber (Zoflex, CD45.1-6S-2) with flat alligator
clips, and the electrical signals were taken to the Source-Me-
ADVANCED ENGINEERING MATERIALS 2008, 10, No. 4
Typical profiles of charge and discharge capacities measured on the first energy cell during the first three baseline
cycles are shown in Figure 6. The average capacities measured for two energy cells are (0.3595 ± 0.0039) mAh and
(0.3815 ± 0.0048) mAh respectively, based on a total of 3 measurements per charge or discharge, respectively. The small
difference observed between the charge and discharge capacities indicates that only a small amount of energy loss or
hysteresis occurred. This energy dissipation is due to electrochemical losses inside the energy cell involving internal resistance and chemical changes, and it is transformed into heat
and other losses.
The results of the pressure cycle experiments from the first
specimen are shown in Figure 7. The all-solid-state thin-film
lithium energy cell was charged and discharged under constant uniaxial pressure at each incremental pressure step,
while holding that pressure uniformly constant throughout
the entire charge/discharge cycle. The data is plotted twice in
different forms for clarity, both bar and line form. While the
bar graph is better at depicting the contrast between the
© 2008 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
http://www.aem-journal.com
395
COMMUNICATIONS
Applied Deflection
Loading
COMMUNICATIONS
Pereira et al./Performance of Thin-Film Lithium Energy Cells under Uniaxial Pressure
when it failed to charge. Visual observation revealed that the
top mica substrate of the energy cell was separated from the
Surlyn sealant and peeled off, exposing the underlying layer
of lithium-metal to the ambient air, thus leading to its rapid
oxidation. The charge failure can be explained by the structural failure of the sealant. This structural failure occurred
suddenly above a certain pressure, which indicates that the
sealant layer was squeezed out between the lithium layer and
the mica substrate once the pressure was sufficiently high for
the sealant to fail. Any broken seal allowing air into the
device will cause rapid deterioration of the active layers
leading to charge failure.[21] The lithium-metal anode of this
energy cell with a uniform shiny silvery color turned almost
completely black within seconds, as a result of oxidation
(Fig. 8(a), 8(b)) from contact with ambient air, and the energy
cell did not take any further charges. The compressive
strength of Lipon is typically several hundred MPa. The presFig. 6. Baseline for three charge/discharge capacities with no pressure for specimen
06030006 (left bar is charge, right bar is discharge). Average charge:
sures used in this study are well below the critical pressure of
(0.3665 ± 0.0044) mAh; average discharge: (0.3595 ± 0.0039) mAh.
Lipon and will not affect the charge/discharge performance
of the energy cell. Under very high pressures, the failure of
charge/discharge capacities, the line graph shows the hysterthe battery could be caused by cracks in the Lipon microstruesis in the charge/discharge cycles more clearly.
ture. However, the stresses in this study remained relatively
The results of the uniaxial pressure charge/discharge cycle
low. Other parts of the battery such as the Surlyn sealant that
experiments in Figure 7 indicate that the capacity of this specsqueezed out of the energy cell edges proved to be structuralimen decayed from the baseline level established at zero presly more critical than the Lipon itself.
sure to a fairly stable plateau with a mild slope. Up until a
Another energy cell (06060015) was tested to confirm the
uniaxial pressure of about 0.8 MPa was reached, the charge/
test results discussed heretofore. Typical charge and disdischarge capacities within this plateau are well within the
charge capacities measured under no mechanical solicitations
data deviation regardless of the amount of pressure applied.
for the second energy cell during the first three baseline
After a uniaxial pressure of 0.83 MPa was reached, capacities
cycles are shown in Figure 9. The average charge and disstarted to drop noticeably until failure occurred at 2.8 MPa,
charge capacities measured were (0.3930 ± 0.0011) mAh and
(0.3815 ± 0.0048) mAh, respectively, based
on a total of 3 measurements per charge or
discharge, respectively.
Figures 10 (a) and (b) show the charge
and discharge cycles with an incremental
uniaxial pressure increase of 70 kPa for the
second energy cell. Up to about 1.5 MPa of
uniaxial pressure, the charge/discharge capacities are within the data scatter, and there
was no visible sign of energy cell degradation. Above 1.5 MPa uniaxial pressure, there
is a slight decay in energy cell capacity up to
2.0 MPA. Above the latter pressure, the
decay is more pronounced and energy cell
failure occurs eventually at 2.3 MPa pressure. A similar, sudden failure above a high
pressure was also observed, indicating that
the sealant layer was squeezed out between
the lithium layer and the mica substrate.
Figure 10(b) shows the same data as Figure 10(a) plotted in line mode to illustrate
hysteresis more clearly.
Fig. 7. Charge/discharge capacity results for sample 06030006: a) Charge (left bar) and discharge (right bar) at
Figure 11(a) shows the charge cycle for
each pressure increment; b) Plotted in line mode to illustrate hysteresis. Average charge:
0.2635 ± 0.0398 mAh; average discharge: 0.2467 ± 0.0715 mAh.
one energy cell under the first uniaxial pres-
396
http://www.aem-journal.com
© 2008 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
ADVANCED ENGINEERING MATERIALS 2008, 10, No. 4
Pereira et al./Performance of Thin-Film Lithium Energy Cells under Uniaxial Pressure
COMMUNICATIONS
Fig. 9. Baseline for three charge/discharge capacities with no deflection for specimen
06060015 (left bar is charge, right bar is discharge). Average charge:
0.3930 ± 0.011 mAh; average discharge: 0.3815 ± 0.0048 mAh.
Fig. 8. Substrate separation due to sealant structural failure for sample 06030006: a)
Bottom substrate with thin-film deposition layers on top, showing lithium-metal oxidized topmost layer b) Top substrate seen from the bottom side that was in contact with
the lithium-metal top layer shown in a). Remnants of the sealant layer can be clearly
seen.
sure increment to about 70 kPa immediately after the baseline
testing. As soon as a constant charge of 1 mA starts at time
zero and at 4.2 VDC, the energy cell voltage reads 3.904 VDC
and then slowly rises to the charge voltage in about 700 s.
The current then starts to drop quickly, and the specimen
was considered fully charged when it reached 50 lA in about
1,200 s. The resulting capacity of the device was calculated to
be 0.3783 mAh. Figure 11(b) shows the discharge immediately following the preceding charge at the same pressure,
with an initial voltage of 4.190 VDC, which is slightly lower
than the charge voltage. At a constant discharge current of
1 mA, the voltage drops at a fairly constant rate below
4 VDC. When it reaches about 3.8 VDC in 1,200 s, the voltage
quickly drops to 3.0 VDC, and the specimen is considered to
be fully discharged. The resulting discharge capacity was calculated to be 0.3722 mAh. Figures 11(c) and (d) show a very
similar charge and discharge behavior for the same specimen
when loaded to about 1 MPa uniaxial pressure. The resulting
ADVANCED ENGINEERING MATERIALS 2008, 10, No. 4
charge capacity was 0.3769 mAh and the discharge capacity
was 0.3722 mAh, respectively. Figures 11 (e) and (f) show the
charge/discharge pattern for the same energy cell when
loaded to about 2.2 MPa. At this ending pressure, the charge
time was longer than the 1,200 s shown in the preceding diagrams, which resulted in a charge capacity of 0.3182 mAh
lower than that of the 0.3930 mAh charge baseline. Also at
the same uniaxial pressure, the shorter discharge time of
about 1,000 s resulted in a discharge capacity of 0.3038 mAh,
which is also lower than that of the 0.3815 mAh earlier obtained discharge baseline. Both charge and discharge values
are substantially outside the data scatter range. The failure
occurred in a mechanism similar to that shown in Figures 8
(a) and (b).
Conclusions
The objective of this study was two-fold. The first objective
was to determine if the all-solid-state thin-film lithium energy
cells could withstand the minimal 550 kPa uniaxial pressure
required for composite manufacturing, which both specimens
successfully did. The second objective was to determine the
upper boundary uniaxial pressure limit of operation for the
all-solid-state thin-film lithium energy cells. The two all-solid-state thin-film lithium energy cells tested in the present
study under uniaxial pressure performed well even when
subjected to uniaxial pressures up to about 2.0 MPa. However, pressures higher than this value led to their degradation. The observed degradation was due to the mechanical
failure of the sealant. Above this pressure, the sealant was
squeezed out of the space between the two mica substrates
and the lithium-metal anode layer, which in turn allowed the
ambient air to penetrate into the energy cell core, thus leading
to the rapid degradation of the charge and discharge performance and the ultimate demise of the energy cell. We found
out that, within the observed range, uniformly distributed
© 2008 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
http://www.aem-journal.com
397
COMMUNICATIONS
Pereira et al./Performance of Thin-Film Lithium Energy Cells under Uniaxial Pressure
Fig. 10. Charge/discharge capacity results for sample 06060015, from 0-2.23 MPa pressures: a) Charge (left
bar) and discharge (right bar) at each pressure increment; b) Plotted in line mode to illustrate hysteresis. Average charge 0.3618 ± 0.0686 mAh; average discharge: 0.3521 ± 0.0684 mAh.
packaging characteristics, we found that allsolid-state thin-film energy cells charge/discharge cycles under upwardly increasing
uniform uniaxial pressure are extraordinarily robust and resilient to the effects of uniaxial, uniformly distributed pressure and
constant power charge/sink of 1mAh. If the
overall structure of the energy cell is mechanically robust, i.e., of high structural integrity, the maximum pressure that can be
imposed is expected to be much higher than
the maximum values noted earlier.
The present study indicates that all-solidstate thin-film energy cells can be used as an
integral part of a load-bearing multifunctional, smart material structure if their packaging is of sufficiently high structural integrity. Hence, the goal of using fiber reinforced
laminated composites as the packaging
material for all-solid-state thin-film batteries
in multifunctional smart materials structures
is well within reach.
Received: August 25, 2007
Final version: December 03, 2007
uniaxial pressure had little or no effect on the charge/discharge performance of the all-solid-state thin-film lithium energy cells. Other power charge/draws outside of 1 mAh were
not of interest in this study for the reasons already pointed
out, albeit that they may be considered for future studies.
Apart from other considerations for failure due to the current
Fig. 11. Charge/discharge behaviour for specimen 06060015: a) First charge starts at 70 kPa pressure; b) followed by discharge at the same pressure; c) charge at 1.0 MPa pressure;
d) discharge at the same pressure; e) charge ending at 2.2 MPa pressure; f) discharge ending at the same pressure.
398
http://www.aem-journal.com
© 2008 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
ADVANCED ENGINEERING MATERIALS 2008, 10, No. 4
Pereira et al./Performance of Thin-Film Lithium Energy Cells under Uniaxial Pressure
[1] A. S. Aricò. P. Bruce, B. Scrosati, J. M. Tarascon, W. van
Schlkwijk, Nat. Mater. 2005, 4, 366.
[2] W. Wakihara, O. Yamamoto Lithium Ion Batteries – Fundamentals and Performance, Kodansha-Wiley-VCH,
Weinhein 1998.
[3] X. Yu, J. B. Bates, G. E. Jellison, F. X. Hart, J. Electrochem.
Soc. 1997, 144, 524.
[4] S. D. Jones, J. R. Akridge, J. Power Sources 1995, 54, 63.
[5] S. D. Jones, J. R. Akridge, Solid State Ionics 1996, 86/88,
1291.
[6] S. J. Lee, H. K. Baik, S. M. Lee, Electrochem. Commun.
2003, 5, 32.
[7] J. B. Bates, N. J. Dudney, G. R. Gruzalski, R. A. Zuhr,
A. Choudhury, C. F. Luck, J. Power Sources 1993, 43–44,
103.
[8] J. B. Bates, N. J. Dudney, D. C. Lubben, G. R. Gruzalski,
B. S. Kwak, X. Yu, R. A. Zuhr, J. Power Sources 1995, 54,
58.
[9] B. J. Neudecker, N. J. Dudney, J. B. Bates, J. Electrochem.
Soc. 2000, 147, 517.
[10] J. B. Bates, N. J. Dudney, B. J. Neudecker, B. Wang, New
Trends in Electrochem. Technol: Energy Storage Syst. in
Electron. Gordon and Breach, 2000.
[11] J. Schwenzel, V. Thangadurai, W. Weppner, J. Power
Sources 2006, 154, 232.
[12] J. P. Thomas, M. A. Qidwai, Acta Mater. 2004, 52, 2155.
[13] M. A. Qidwai, J. P. Thomas, P. Matic, Proc. SPIE 2002,
4698, 180.
[14] G. Nagasubramanian, D. H. Doughty, J. Power Sources
2004, 136, 395.
[15] E. P. Roth, D. H. Doughty, J. Power Sources 2004, 128,
308.
[16] E. V. Thomas, H. L. Case, D. G. Doughty, R. G. Jungst,
G. Nagasubramanian. E. P. Roth, J. Power Sources 2003,
124, 254.
[17] D. H. Doughty, E. P. Roth, C. C. Crafts, G. Nagasubramanian, G. Henriksen, K. Amine, J. Power Sources 2005,
146, 116.
[18] D. H. Doughty, P. C. Butler, R. G. Jungst, E. P. Roth,
J. Power Sources 2002, 110, 357.
[19] T. Pereira, R. Scaffaro, S. Nieh, J. Arias, Z. Guo, H. T.
Hahn. J. Micromech. Microeng. 2006, 16, 2714.
[20] V. Krasnov, K.-W. Nieh, S.-J. Ting. Thin Film Battery and
Meth. of Manufact. U.S. Patent No. 6,632, 563, http://
www.uspto.gov/.
[21] W. C. West, J. F. Whitacre, V. WHite, B. V. Ratnakumar,
J. Micromech. Microeng. 2002, 12, 58.
[22] J. W. Long, B. Dunn, D. R. Rolison, H. S. White, Chem.
Rev. 2004, 104, 4463.
[23] F. Croce, F. S. Fiory, L. Persi, B. Scrosati, Electrochem.
Solid-State Lett. 2001, 4 (8), A121.
______________________
ADVANCED ENGINEERING MATERIALS 2008, 10, No. 4
© 2008 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
http://www.aem-journal.com
399
COMMUNICATIONS
–